nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 31 Packages Projects Releases Wiki Activity

chore : correct typos [no ci] (#20041)

Browse Source 
* fix(docs): correct typos found during code review

Non-functional changes only:
- Fixed minor spelling mistakes in comments
- Corrected typos in user-facing strings
- No variables, logic, or functional code was modified.

Signed-off-by: Marcel Petrick <mail@marcelpetrick.it>

* Update docs/backend/CANN.md

Co-authored-by: Aaron Teo <taronaeo@gmail.com>

* Revert "Auxiliary commit to revert individual files from 846d1c301281178efbc6ce6060ad34c1ebe45af8"

This reverts commit 02fcf0c7db661d5ff3eff96b2b2db9fdb7213256.

* Update tests/test-backend-ops.cpp

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

* Update tests/test-backend-ops.cpp

Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>

---------

Signed-off-by: Marcel Petrick <mail@marcelpetrick.it>
Co-authored-by: Aaron Teo <taronaeo@gmail.com>
Co-authored-by: Sigbjørn Skjæret <sigbjorn.skjaeret@scala.com>
This commit is contained in:
Marcel Petrick
2026-03-05 08:50:21 +01:00
committed by GitHub
co-authored by Aaron Teo Sigbjørn Skjæret
parent 7a99dc85e2
commit 92f7da00b4
81 changed files with 160 additions and 160 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-hexagon/ggml-hexagon.cpp
+11 -11
View File
@@ -139,7 +139,7 @@ struct ggml_hexagon_session {
};
void ggml_hexagon_session::enqueue(struct htp_general_req &req, struct dspqueue_buffer *bufs, uint32_t n_bufs, bool sync) {
// Bump pending flag (cleared in the session::flush once we get the responce)
// Bump pending flag (cleared in the session::flush once we get the response)
this->op_pending++; // atomic inc
int err = dspqueue_write(this->queue,
@@ -443,7 +443,7 @@ static void repack_row_q4x4x2(uint8_t * y, const block_q4_0 * x, int64_t k) {
// Repack the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_Q4_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Repack the scales
ggml_half * d = (ggml_half *) (y_d + i * dblk_size);
@@ -503,7 +503,7 @@ static void unpack_row_q4x4x2(block_q4_0 * x, const uint8_t * y, int64_t k) {
// Repack the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_Q4_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Unpack the scales
const ggml_half * d = (const ggml_half *) (y_d + i * dblk_size);
@@ -552,7 +552,7 @@ static void init_row_q4x4x2(block_q4_0 * x, int64_t k) {
// Init the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_Q4_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Unpack the scales
x[i * 8 + 0].d = 0;
@@ -770,7 +770,7 @@ static void repack_row_q8x4x2(uint8_t * y, const block_q8_0 * x, int64_t k) {
// Repack the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_Q4_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Repack the scales
ggml_half * d = (ggml_half *) (y_d + i * dblk_size);
@@ -829,7 +829,7 @@ static void unpack_row_q8x4x2(block_q8_0 * x, const uint8_t * y, int64_t k) {
// Repack the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_Q4_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Unpack the scales
const ggml_half * d = (const ggml_half *) (y_d + i * dblk_size);
@@ -878,7 +878,7 @@ static void init_row_q8x4x2(block_q8_0 * x, int64_t k) {
// Init the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_Q8_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Unpack the scales
x[i * 8 + 0].d = 0;
@@ -1120,7 +1120,7 @@ static void repack_row_mxfp4x4x2(uint8_t * y, const block_mxfp4 * x, int64_t k)
// Repack the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_MXFP4x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Repack the scales
uint8_t * e = (uint8_t *) (y_e + i * eblk_size);
@@ -1180,7 +1180,7 @@ static void unpack_row_mxfp4x4x2(block_mxfp4 * x, const uint8_t * y, int64_t k)
// Repack the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_MXFP4_0x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Unpack the scales
const uint8_t * e = (const uint8_t *) (y_e + i * eblk_size);
@@ -1229,7 +1229,7 @@ static void init_row_mxfp4x4x2(block_mxfp4 * x, int64_t k) {
// Init the scales
// Note: Do not combine with the loop above. For tensor sizes not multiple of 256 (QK_MXFP4x4x2)
// the last block is truncated and overriden by the scales.
// the last block is truncated and overridden by the scales.
for (int i = 0; i < nb; i++) {
// Unpack the scales
x[i * 8 + 0].e = 0;
@@ -2670,7 +2670,7 @@ static std::vector<int> ggml_hexagon_graph_optimize_reorder(const std::vector<no
// The main goal here is to stack the MUL_MAT ops with the same src1 input.
// This allows use to reuse dynamically quantized src1 in VTCM.
// TODO: the current version might do incorrect reodering in cases where quantized src0
// TODO: the current version might do incorrect reordering in cases where quantized src0
// input is an output of another Op.
for (int i0 = 0; i0 < n; i0++) {
ggml/src/ggml-hexagon/htp-drv.cpp
+1 -1
View File
@@ -282,7 +282,7 @@ static std::string get_driver_path() {
// Replace \SystemRoot with an absolute path from system ENV windir
const std::wstring systemRootEnv = L"windir";
// Query the number of wide charactors this variable requires
// Query the number of wide characters this variable requires
DWORD numWords = GetEnvironmentVariableW(systemRootEnv.c_str(), NULL, 0);
if (numWords == 0) {
GGML_LOG_ERROR("ggml-hex: Failed get systemRoot environment variable\n");
ggml/src/ggml-hexagon/htp/hvx-inverse.h
+1 -1
View File
@@ -67,7 +67,7 @@ static inline HVX_Vector hvx_vec_inverse_f16(HVX_Vector vals) {
HVX_Vector vcl0 = Q6_Vuh_vcl0_Vuh(rm); //count leading zeros
// Get mantissa for 16-bit represenation
// Get mantissa for 16-bit representation
HVX_Vector mant_recip = Q6_V_vand_VV(Q6_Vh_vasr_VhR(Q6_Vh_vasl_VhVh(rm, vcl0), 5), Q6_Vh_vsplat_R(0x03FF));
//Compute Reciprocal Exponent
ggml/src/ggml-hexagon/htp/rope-ops.c
+1 -1
View File
@@ -18,7 +18,7 @@
#include "htp-msg.h"
#include "htp-ops.h"
// Redefined the types GGML_ROPE_TYPE_NORMAL & GGML_ROPE_TYPE_NEOX as we cant include ggml.h
// Redefined the types GGML_ROPE_TYPE_NORMAL & GGML_ROPE_TYPE_NEOX as we can't include ggml.h
#define HTP_ROPE_TYPE_NORMAL 0
#define HTP_ROPE_TYPE_NEOX 2
ggml/src/ggml-hexagon/htp/worker-pool.c
+1 -1
View File
@@ -56,7 +56,7 @@ static void worker_pool_main(void * context) {
unsigned int n = atomic_load(&pool->n_jobs);
unsigned int i = atomic_fetch_add(&pool->next_job, 1);
if (i >= n) {
// Spurios wakeup
// Spurious wakeup
continue;
}